Placeto: Learning Generalizable Device Placement Algorithms for Distributed Machine Learning-猫眼课题宝

DOI：

发表时间：

2019-06

期刊：

影响因子：

通讯作者：

中科院分区：

其他

文献类型：

作者：

Ravichandra Addanki;S. Venkatakrishnan;Shreyan Gupta;Hongzi Mao;Mohammad Alizadeh

我们提出了Placeto，这是一种强化学习（RL）方法，可以有效地为分布式神经网络训练找到设备位置。与之前只为特定计算图找到设备放置的方法不同，Placeto可以学习可应用于任何图的可推广的设备放置策略。在我们的方法中，我们提出了两个关键思想：（1）我们将策略表示为执行迭代放置改进，而不是一次性输出放置;（2）我们使用图嵌入来捕获有关计算图结构的相关信息，而不依赖于节点标签进行索引。这些想法使Placeto能够有效地训练并推广到看不见的图形。我们的实验表明，Placeto需要少6.1倍的训练步骤来找到与先前方法找到的最佳位置相同或更好的位置。此外，Placeto能够为任何给定的图族学习可推广的放置策略，然后可以在没有任何重新训练的情况下使用该策略来预测来自同一族的未见过图的优化放置。这消除了先前RL方法所产生的巨大开销，这些方法缺乏可推广性，每次放置新图时都需要从头开始重新训练。

We present Placeto, a reinforcement learning (RL) approach to efficiently find device placements for distributed neural network training. Unlike prior approaches that only find a device placement for a specific computation graph, Placeto can learn generalizable device placement policies that can be applied to any graph. We propose two key ideas in our approach: (1) we represent the policy as performing iterative placement improvements, rather than outputting a placement in one shot; (2) we use graph embeddings to capture relevant information about the structure of the computation graph, without relying on node labels for indexing. These ideas allow Placeto to train efficiently and generalize to unseen graphs. Our experiments show that Placeto requires up to 6.1x fewer training steps to find placements that are on par with or better than the best placements found by prior approaches. Moreover, Placeto is able to learn a generalizable placement policy for any given family of graphs, which can then be used without any retraining to predict optimized placements for unseen graphs from the same family. This eliminates the large overhead incurred by prior RL approaches whose lack of generalizability necessitates re-training from scratch every time a new graph is to be placed.

Placeto: Learning Generalizable Device Placement Algorithms for Distributed Machine Learning

文献摘要

相似文献